Company history

Every Meiji-era attempt to industrialise flat glass in Japan had failed on the technology, so whether a company could obtain a machine process decided whether it had a business at all. Sugita Yosaburo, a Chicago graduate working for an Osaka trading house, happened in 1914 to see the Colburn drawing process being industrialised in Toledo, and changed trades on the spot. Returning to America in 1917 he sought help from Sumitomo’s New York manager, who replied that a business good for the country deserved backing. Libbey-Owens would not licence below ¥4m — impossible in cash for a company capitalised at ¥3m, half paid in — so the patent was contributed in kind, and the American firm took 20,000 shares, a third of the company, plus $100,000. Nichibei Sheet Glass — “Japan-America Sheet Glass” — was incorporated in Osaka in November 1918, its parentage written into its name.

The Colburn machine arrived at Futajima in Fukuoka in November 1919 with Libbey engineers to install it. Drawing began at midnight on 1 October 1920; by eight in the morning the cloudiness had cleared and glass about three millimetres thick was running steadily. It was the first sheet glass in Japan made by automatic continuous drawing — the end of hand-blown production.

The postwar slump arrived almost immediately. In August 1922 an extraordinary general meeting handed money and management alike to Sumitomo, and the pattern was set: technology borrowed from America, capital and management borrowed from Sumitomo. The name became Nippon Sheet Glass in 1931, though Libbey kept a director on the board until 1938. War severed the connection and put the American holding under enemy-property administration; when the unpaid share capital was called in July 1949, Libbey paid — the first inward foreign investment in postwar Japan — and its shares were restored in full.

The shares listed in Tokyo, Osaka and Kobe in June 1950, and capital multiplied sixteen-fold in five years to fund reconstruction demand. Two structural facts shaped everything after. First, flat glass is a product whose maker cannot be told by looking, so left alone it falls into price competition; differentiation means heat-absorbing or laminated functions, and that means the absolute size of the R&D budget decides. Second, the Japanese industry had been founded by two companies — Asahi Glass in 1907 and this one in 1918 — and the smaller of them would spend ninety years fighting a scale contest at a scale disadvantage. In 1954 it set up a subsidiary for automotive glass, a second pillar beside construction; in fiscal 2025 that pillar was the larger of the two.

Then the process changed once and for all. Alastair Pilkington conceived of floating molten glass on molten tin in 1952 — the idea came while washing dishes — and after seven years and some $11.1M (¥4bn) of research his company announced perfect float glass in 1959. One week after the announcement, Nippon Sheet Glass approached Pilkington about a licence. Its president visited in 1961, his successor negotiated in London in 1963, and the contract was signed in March 1964. Pilkington was reportedly startled by how short and how concrete the Japanese construction plan was compared with the European and American licensees that had gone first.

On 15 November 1965, in the snow at Maizuru, the tap was opened and molten glass drew into the float bath, emerging some forty minutes later: the first float glass in Asia. The borrowed process had been rebuilt for where it stood — earthquake provisions the British never had to consider, continuous turning and branching equipment after the cutting line because the speed was so high, and a newly developed air table that would not scratch the surface. Float lines followed at Chiba in 1971 and Maizuru in 1978. Abroad, however, the company put down only one significant marker, a joint venture in Malaysia in 1971, followed by an automotive venture in Mexico in 1975 — a thinness that would later be cited, again and again, as the reason for the takeover.

In 1986 the company formed an American automotive-glass venture with Libbey-Owens-Ford — a Pilkington subsidiary, and the successor of the very firm whose patent had created it in 1918 — and took 20% of LOF in 1990. Sixty-eight years on, the licensor and licensee were tied by capital as well as technology. At home the 1990s were a decade of tidying: fibre operations sold, plants at Kawasaki and elsewhere closed, subsidiaries absorbed or exchanged away.

In 2000 Nippon Sheet Glass bought 10% of Pilkington, and in October 2001 raised it to 20% to become the largest shareholder, accounting for it by the equity method. The logic — complementary technology, complementary geography, a flat-glass market growing a point faster than GDP — was clear on paper. Five years later the result was not. Pilkington described the relationship as close but not special; joint purchasing does not work between separate companies, and neither side disclosed all of its research, because each calculated that handing over technology it might lose was a poor trade. A 20% holder was still a competitor.

Meanwhile the businesses meant to sit outside flat glass failed to grow. Information and electronics — optical-communication lenses above all — bled after the dot-com collapse; the LCD substrate business was a joint venture with HOYA holding around a tenth of the market and could not be moved unilaterally. Asahi Glass earned roughly half its 2004 operating profit in electronics and displays; Nippon Sheet Glass missed that wave entirely. Revenue slipped from $2.4B (¥287bn) in the year to March 2002 to $2.4B (¥266bn) four years later. The year before it made its offer, it was a solid domestic number two with no next move.

Why not simply stay at 20%? Because British takeover rules require a bidder crossing 30% to offer for the whole company. Chairman Izuhara Yozo said that on capital efficiency alone, 51% would have done; the rule closed the path of buying up gradually to fit one’s means. The choice was 20% or 100%. Nippon Sheet Glass announced an offer at 150 pence in November 2005, was refused, and settled in February 2006 at 165 pence — a 30% premium. The acquisition completed that June at about $3.3B (¥359bn) for the equity and roughly $5.6B (¥616bn) in all, swallowing a company nearly twice its size. Revenue went from $2.4B (¥266bn) to $7.3B (¥866bn) in two years, employees from 12,736 to 35,811, and overseas sales from a fifth to nearly four-fifths of the total.

The financing was the problem. About $3.3B (¥363bn) came from bank borrowing and $998.6M (¥110bn) from convertible bonds; interest-bearing debt went from $1.1B (¥125bn) to over $5.2B (¥600bn), and the equity ratio fell from 48.7% to 23.9%. Goodwill of roughly $1.9B (¥220bn) would depress operating profit by about $94.6M (¥11bn) a year even amortised over twenty years, against synergies the company itself put at $47M (¥4bn) annually — less than half the amortisation, disclosed at the time of the deal. Then Lehman hit, Europe was half of revenue, and the losses began.

Governance followed the balance of the group. With most directors and staff coming from Pilkington and Japanese career employees a fifth of the whole, the company promoted Stuart Chambers, from the acquired side, to president in 2008 and moved to a committee-based board; Chambers left after little more than a year for family reasons, and an outside American appointee, Craig Naylor, resigned in 2012 over strategy. Three presidents in four years reflected less the individuals than the fact that a company with one overseas plant before the deal had never built managers who could run one with plants in twenty-eight countries. Restructuring took out 6,700 jobs by March 2011 and 3,500 more in 2012; Moody’s cut the rating to Ba3; the dividend was suspended for the first time in thirty-five years in 2013.

In the second quarter of fiscal 2022 the company impaired $371.5M (¥49bn) of the goodwill and intangibles created in 2006, in European automotive glass — its own record that the price paid could not be recovered from the earnings bought. European Commission cartel fines inherited with Pilkington had already cost $653.2M (¥68bn). Ten of the nineteen years after the acquisition ended in net loss, $994.1M (¥151bn) in aggregate. With interest-bearing debt of $3.8B (¥570bn) at the end of 2025, $1.2B (¥177bn) of it due within a year, the company announced in March 2026 a recapitalisation: $1.1B (¥165bn) from Apollo Global Management for 72% of the votes, $935.5M (¥140bn) of bank debt converted to equity, debt cut by $1.5B (¥231bn), minority shareholders paid ¥500 a share, and delisting scheduled for November 2026. A company begun with an American patent and Sumitomo’s money leaves the market, 108 years later, in the hands of an American fund.